• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 479 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMR33630: AVERAGE model isn't posted, but Webench computes the bode plot, can I still get a copy of the average model?

Andre Bezanson
Prodigy 10 points
Part Number: LMR33630
Other Parts Discussed in Thread: TINA-TI, ,

I've noticed that when running Webench on the LMR33630ADDAR I can get it to generate an average model that gives a good estimate of the gain/phase margin. However, I can only download the transient spices models on the product page and the Tina-Ti export isn't available from Webench. 

Would someone from the TI team be willing to give me a copy of the Average model used for the simulation? 
OR
Can someone shed some light on why this isn't an option? 

  • 0
    TI__Guru 73545 points

    Hello

    You can use the attached information to build a model.

    Please keep in mind that this information is proprietary to TI.

    Also, normal variations in device parameters and temperature will cause differences in results between

    bench and simulations.

    Thanks

    LMR33630_loop.pptx

  • 0
    Prodigy 10 points

    Thanks for the response, I took a shot at building out a model using that data, but my understanding is limited. So far my results don't match and are shown in the below image.
    Any suggestions on how I can improve my model? (Or is there a reference model I could build from?)  

  • 0 in reply to Andre Bezanson
    TI__Guru 73545 points

    Hello

    Sorry for the confusion.

    Please set up your simulation as shown in the attached diagram using the 

    parameter values shown in the diagram.

    Thanks

  • 0 in reply to Frank De Stasi
    Prodigy 10 points

    Thanks for the updated information, I think it's helped a bit, as the simulation does regulate to 5V when running a transient simulation. However, I still can't get a meaningful AC analysis of the circuit. 

    I've setup the circuit with the intention of mirroring the LMR33630ARNXEVM eval board so that I can double check the results. 

    As you can see in the image, my simulation is still off when it comes to gain and phase. Any tips on how I can further refine this? 

  • 0 in reply to Andre Bezanson
    TI__Guru 73545 points

    Hello

    You miscalculated R4.  It should be 5.5 Ohms.

    Thanks

  • 0 in reply to Frank De Stasi
    Prodigy 10 points

    With the model updated to use the correct Ro value, and with the AC signal source positioned like it is in the Webench design tool I get this result: 
    - The gain seems about right, but the phase is far off from the predicted and measured values (see attached image).

    Any suggestions on how to get the phase data to be more precise? Or is that just a limitation of the first order model we've been working from?

    - The reason why this model is critical to our work is because we need a part that can achieve ultra-low EMI for our design. The use of input and output filters is critical to meeting the design criteria, however, we need a solid model to build off of. We are a bit torn, as the part has very good performance, and is ideal for our use case, however, we need this model to build from as we build out the filters. Other parts with the average model don't work quite as well for us.  

    https://imgur.com/a/p3bEREK

    Would it be possible to get the model that is used directly by WeBench? We'd be happy to sign any sort of NDA or release

  • 0 in reply to Andre Bezanson
    TI__Guru 73545 points

    Hello

    Can you tell me how close your model is to Webench ??

    I will talk to our Webench team and see if the average model is available.

    Thanks

  • 0 in reply to Frank De Stasi
    TI__Guru 73545 points

    Hello

    Thank you for your patience.

    I have contacted our Webench team and they are working to get the average model.

    Thanks

  • +1 in reply to Frank De Stasi
    TI__Guru 73545 points

    Hello

    I have attached the model here.

    And the response from our Webench team.

    I hope this helps.

    Thanks

    Sorry for the delay. I have attached average netlist model of LMR33630-SOIC (LMR33630A-SOIC_AVG.txt).

    Please note that the average model  used on WEBENCH is a netlist and it is derived from the template. Since this an IC netlist,  you need to configure the external component manually. For ease of use , I have also  attached LMR33630_Pspice.cir file which can be run directly into AMS simulator. Please replace the external component as per your requirement.

    LMR33630_PSpice.cir

    Fullscreen LMR33630A-SOIC_AVG.txt Download 
    <html><head><meta http-equiv="Content-Type" content="text/html; charset=utf-8"/><meta name="decorator" content="none"/><script src="/s/-151011903/c3bc80a/wv5i03/4.2.17/_/download/resources/com.atlassian.plugins.authentication.atlassian-authentication-plugin:save-fragment/js.cookie.js"></script><script>if(window.location.hash) {var expirationDate = new Date(); expirationDate.setTime(expirationDate.getTime() + 60000 * 5); Cookies.set('atlassian-authentication-plugin-url-fragment_b1084dbb-c68f-4c77-a052-ec9c1dc4b761', window.location.hash.substr(1), {expires : expirationDate, path : '', secure: true, SameSite: 'None' });}window.location.assign("https://entlogin.ti.com/idp/SSO.saml2?SAMLRequest=fZHLbsIwEEV/JfLecQhpwBYg0VK1SJQgknbRTeWEKVhN7DQzQf38hpdKNyzHnnN1rmaEuiprNW1pZ9fw3QKS91OVFtXxY8zaxiqn0aCyugJUVKh0%2BrJQoR%2BounHkCleyK%2BQ2oRGhIeMs8%2BazMUuWj4vkab782MhhLmUkuY7zgEcyHnANYcj1Jwwh7steHAXMe4MGO3bMuqguALGFuUXSlrqnIOzz4I73ZBZEqjdU/cE782ZdH2M1HakdUY1KCLBUuq2xPhm/cJUwm1qkaeIf9EPmrc6t7o3dGLu9XSg/LaF6zrIVXyVpxrzppeSDs9hW0KTQ7E0Br%2BvFn0RuKG%2BLLyDf0PZiUpdt54Wi4/clkDgYFecQNhkdRnWs3Uxu5YzE9eZp%2Bn/jyS8%3D&RelayState=b1084dbb-c68f-4c77-a052-ec9c1dc4b761");</script></head><body></body></html>

  • 0 in reply to Frank De Stasi
    Prodigy 10 points

    Thanks for the model! That works well.

    It took a little bit of effort to make it work in my spice software (LTSpice and Tina), as I needed to:
    - to split out the reference model from the actual model, (starting at line 214) to get the model to run
    - to make sure the model used the same ESR/ESL and capacitor derating values to get the same phase/gain margin values as the reference design.

    From there, it worked well and has been a huge help in our design work. Thank you!

  • 0 in reply to Andre Bezanson
    TI__Guru 73545 points

    Hello

    Glad we could help.

    Thanks